TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Earliest Eligible Virtual Deadline First: Flexible Resource Allocation (1995)

17 pointsby pantalaimonover 1 year ago

3 comments

tux3over 1 year ago
For context, EEVDF is the new scheduler in Linux 6.6: <a href="https:&#x2F;&#x2F;kernelnewbies.org&#x2F;Linux_6.6#New_task_scheduler:_EEVDF" rel="nofollow noreferrer">https:&#x2F;&#x2F;kernelnewbies.org&#x2F;Linux_6.6#New_task_scheduler:_EEVD...</a>
samsquireover 1 year ago
Just a warning I am a beginner on scheduling. Just reading about events in this paper, I assume that&#x27;s IO events?<p>I have a multithreaded ringbuffer and I&#x27;ve been trying to work out how to reduce latency with Linux scheduling and thread core affinity. It doesn&#x27;t seem to improve latency<p>I installed Ubuntu&#x27;s realtime kernel and played around with SCHED_DEADLINE but I don&#x27;t think real time kernel is what&#x27;s needed to schedule long running threads continuously ?<p>What I want is the latency between scheduling of two given threads to be as close together as possible. Is this possible ? For reference LMAX Disruptor in Java can send a message between thread with average latency of 52 nanoseconds.<p>My naive toy C implementation is 80-130 nanoseconds from glances.
m463over 1 year ago
is this postscript? also this seems to be (1996)
评论 #38082723 未加载